1. Field of the Invention
The present invention relates to a communication device, a network system using the same, and a method of constructing a spanning tree, and more particularly, to a method of constructing a spanning tree in a network system in which a plurality of networks are interconnected through bridges.
2. Description of the Prior Art
A conventional bridge network using a spanning tree will be described with reference to FIGS. 7 and 10. In the network shown in FIG. 7, bridges 1, 3, and 4 are interconnected by a local area network (LAN) A; bridges 1, 2, and 5 are interconnected by a LAN B; and bridges 2, 3 and 4 are interconnected by a LAN C.
Further, bridges 5, 6, 8, and 9 are interconnected by a LAN D; bridges 7, 8, and 9 are interconnected by a LAN E; and bridges 6 and 7 are interconnected by a LAN F.
Each of the bridges 1 to 9 has the function of constructing a spanning tree by transmitting and receiving configuration-bridge protocol data units (C-BPDU) specified in IEEE802.1d. FIG. 10 shows the state of a spanning tree constructed as a result of exchange of CBPDUs.
A spanning tree is constructed with the bridge 1 determined as a root bridge, as indicated by the dotted line in FIG. 10. A spanning tree is thus constructed to enable transfer of data packets without forming a loop on the network formed of a plurality of LANs connected by bridges.
In FIGS. 7 and 10, “BRIDGE ID” is an identification number for identification of each bridge, and “PORT ID” is an identification number for each port. Also, “ROOT PATH COST” and “PATH COST” are values each representing the facility for reaching to the root bridge. Each time a spanning tree is constructed and a signal passes one of the bridges, the value of the path cost at the bridge (port) is incremented. The root path cost and the path cost are used as parameters such as to enable selection of the bridge (port) with the minimized root path cost (easier to reach).
However, there is a problem with the conventional spanning tree, as described below. When the link between some of the bridges constituting the network is lost due to a fault in the bridges or a fault in the network connected between the bridges, reconstruction of a spanning tree is performed in accordance with a rule specified by IEEE802.1d. When this reconstruction is performed, spanning tree configuration information on the constructed spanning tree and a filtering data base for data packet forwarding obtained by learning are entirely initialized and a tree is newly constructed.
A considerably long time, depending on the number of bridges constituting the reconstructed network of the tree, is required for the reconstruction, and data packets to be transferred onto the network are not transferred until the reconstruction is completed. That is, a condition similar to service interruption occurs. If the scale of the bridge network is larger (the number of bridges constituting the network is larger), a worse condition results.